Influence of surface curvature on light-based nondestructive measurement of stone fruit

Highlights

• Intact stone fruit were modeled as concentric-spherical-layered tissues.

• Light propagation in fruit was simulated using Monte Carlo methods.

• Surface curvature affects the diffuse reflectance and the inspection efficiency.

• The experimental arrangements require adjustment accordingly.

Abstract

The propagation of photons in stone fruit tissue was simulated with the Monte Carlo (MC) method. A model of concentric-spherical-layered tissues was built for intact stone fruit. Peaches were used as typical representatives of stone fruit. Diffuse reflectance and inspection efficiency were calculated using the concentric-spherical-layered model. The simulation results were compared with those expected from the classic infinitely-wide planar layered model. Surface curvature increases the diffuse reflectance but decreases the inspection efficiency, especially for the case of short source-detector distance. For small size fruit items, errors in the planar layered tissues model would be increased due to their greater surface curvature. This paper suggests a need to take into account the influence of surface curvature to retrieve the optical absorption and reduced scattering coefficients from Vis-NIR spectroscopy data for fruit. And the experimental arrangements such as the source-detector distance and the integration time for the spectrometer require adjustment accordingly.

Introduction

Visible-near infrared (Vis-NIR) spectroscopy measurements have been utilized for fruit quality inspection for many years (Nicolaï et al., 2007). The fruit is irradiated with Vis-NIR radiation, and the reflected or transmitted radiation is measured. The radiation interacts with fruit tissue through both absorption and scattering. While absorption in the Vis-NIR range is related to some important chemical quality attributes such as the total sugar content, scattering is related to the microstructure of the fruit (Oey et al., 2007). With the optical properties retrieved from the Vis-NIR spectroscopy measurements, the quality and internal structure of fruit can be deduced by inverse methods such as linear regression techniques and nonlinear regression techniques (Nicolaï et al., 2007). Unfortunately, the light-based techniques of fruit quality inspection are still not well developed, and their accuracy and robustness is often limited. So there is a growing demand for accurate and fast models to quantitatively understand light transport process and features in fruits.

Fruits are a kind of highly scattering turbid media. Although the radiation transport equation (RTE) is the basis of the general solution of light transport in turbid media (Ishimaru, 1978, Schweiger et al., 1995), its application in complex geometrics is hindered by difficulties in obtaining the analytical solution to RTE. Monte Carlo (MC) simulation methods solve the RTE numerically and have successfully tested the validity of analytical algorithms obtained from the RTE (Wang and Jacques, 1992). MC based methods have been widely utilized to ray-trace individual photons in biological medium such as fruits, vegetables, and human tissues (Fraser et al., 2003, Okada et al., 1997, Tsai et al., 2001, Wang and Liang, 1999, Wilson and Adam, 1983). Several studies (Baranyai and Zude, 2009, Fraser et al., 2003, Qin and Lu, 2009) have used MC simulation to investigate light propagation in fruit tissues. However, these simulations consider the fruit to be either an infinitely-wide planar layered turbid media (Baranyai and Zude, 2009, Qin and Lu, 2009), or a homogeneous sphere of tissue (Fraser et al., 2003). Actually, fruits are neither infinitely-wide nor homogeneous. Take stone fruit (or drupe) for example, it is an indehiscent fruit in which an outer fleshy part surrounds a shell of hardened endocarp with a seed inside. The geometries and optical properties of different structures of stone fruit would inevitably affect the light transport features.

In this paper, peaches were used as typical representatives of stone fruit. We developed a MC model of concentric-spherical-layered tissues and compared the simulation results with those obtained by the planar layered tissues model. The influence of surface curvature on light-based nondestructive inspection of stone fruit was assessed. The analysis would result in a better insight into the interaction of light radiation with the fruit tissue, and help the development of optical techniques for fruit quality inspection.